The calculation of actual or received dose intensity: a comparison of published methods

Two recent reports of the same combination chemotherapy program, cyclophosphamide, doxorubicin, etoposide, methotrexate, cytarabine, vincristine, bleomycin, and prednisone (ProMACE-CytaBOM), in similar subsets of patients with advanced-stage aggressive-histology lymphoma used two different methods to report the actual dose-intensity (DI) data. One method treats DI as a property of a particular cycle of treatment within the entire population that received that cycle. The other treats DI as a characteristic of a particular patient's entire treatment course. We have applied both methods to the same set of data and provide evidence that the latter method is preferable for at least two reasons: (1) it more accurately reflects actual DI by clearly incorporating the duration of the actual treatment course and, thus, can be used to compare the administration of same or related regimens to distinct patient populations; and (2) it allows assignment of a numerical value to an individual patient's treatment course or a group of patients' treatment courses such that DI can be examined for its impact on treatment outcome just like any other prognostic factor. The observed differences in treatment outcome between the Southwest Oncology Group (SWOG) and National Cancer Institute (NCI) studies are not clearly related to differences in distribution of clinical prognostic factors in the two study populations. The differences in methods of reporting DI prohibit evaluation of the influence of dose-related variables on outcome in the two studies. Adoption of a standard method of calculating and reporting DI data would facilitate evaluation of the prognostic significance of DI.